Fix broken SVD tests (#33)

* reenable broken tests

* use `isstored` instead of `I in eachstoredindex`

* Bump version and minimal SparseArraysBase compat

* Formatter

Author: lkdvos

Project.toml

@@ -1,7 +1,7 @@
 name = "BlockSparseArrays"
 uuid = "2c9a651f-6452-4ace-a6ac-809f4280fbb4"
 authors = ["ITensor developers <support@itensor.org> and contributors"]
-version = "0.2.10"
+version = "0.2.11"
 
 [deps]
 Adapt = "79e6a3ab-5dfb-504d-930d-738a2a938a0e"
@@ -42,7 +42,7 @@ LabelledNumbers = "0.1.0"
 LinearAlgebra = "1.10"
 MacroTools = "0.5.13"
 MapBroadcast = "0.1.5"
-SparseArraysBase = "0.2.2"
+SparseArraysBase = "0.2.10"
 SplitApplyCombine = "1.2.3"
 TensorAlgebra = "0.1.0"
 Test = "1.10"

src/abstractblocksparsearray/views.jl

@@ -68,11 +68,8 @@ function BlockArrays.viewblock(
   a::AbstractBlockSparseArray{<:Any,N}, block::Vararg{Block{1},N}
 ) where {N}
   I = CartesianIndex(Int.(block))
-  # TODO: Use `eachblockstoredindex`.
-  if I ∈ eachstoredindex(blocks(a))
-    return blocks(a)[I]
-  end
-  return BlockView(a, block)
+  # TODO: isblockstored
+  return isstored(blocks(a), I) ? blocks(a)[I] : BlockView(a, block)
 end
 
 # Specialized code for getting the view of a subblock.

test/test_svd.jl

@@ -5,11 +5,11 @@ using LinearAlgebra: LinearAlgebra
 using Random: Random
 using Test: @inferred, @testset, @test
 
-function test_svd(a, usv; broken=false)
+function test_svd(a, usv)
   U, S, V = usv
-  @test U * diagonal(S) * V' ≈ a broken = broken
-  @test U' * U ≈ LinearAlgebra.I
-  @test V' * V ≈ LinearAlgebra.I
+  return (U * diagonal(S) * V' ≈ a) &&
+         (U' * U ≈ LinearAlgebra.I) &&
+         (V' * V ≈ LinearAlgebra.I)
 end
 
 # regular matrix
@@ -19,7 +19,7 @@ eltypes = (Float32, Float64, ComplexF64)
 @testset "($m, $n) Matrix{$T}" for ((m, n), T) in Iterators.product(sizes, eltypes)
   a = rand(m, n)
   usv = @inferred svd(a)
-  test_svd(a, usv)
+  @test test_svd(a, usv)
 end
 
 # block matrix
@@ -28,7 +28,7 @@ blockszs = (([2, 2], [2, 2]), ([2, 2], [2, 3]), ([2, 2, 1], [2, 3]), ([2, 3], [2
 @testset "($m, $n) BlockMatrix{$T}" for ((m, n), T) in Iterators.product(blockszs, eltypes)
   a = mortar([rand(T, i, j) for i in m, j in n])
   usv = svd(a)
-  test_svd(a, usv)
+  @test test_svd(a, usv)
   @test usv.U isa BlockedMatrix
   @test usv.Vt isa BlockedMatrix
   @test usv.S isa BlockedVector
@@ -39,17 +39,8 @@ end
 @testset "($m, $n) BlockDiagonal{$T}" for ((m, n), T) in
                                           Iterators.product(blockszs, eltypes)
   a = BlockDiagonal([rand(T, i, j) for (i, j) in zip(m, n)])
-  if VERSION ≥ v"1.11"
-    usv = svd(a)
-    # TODO: `BlockDiagonal * Adjoint` errors
-    # TODO: This is broken because of https://github.com/JuliaLang/julia/issues/57034,
-    # fix and reenable.
-    test_svd(a, usv; broken=true)
-  else
-    # `svd(a)` depends on `diagind(::AbstractMatrix, ::IndexStyle)`
-    # being defined, but it was only introduced in Julia v1.11.
-    @test svd(a) broken = true
-  end
+  usv = svd(a)
+  @test test_svd(a, usv)
 end
 
 # blocksparse 
@@ -60,25 +51,25 @@ end
 
   # test empty matrix
   usv_empty = svd(a)
-  test_svd(a, usv_empty)
+  @test test_svd(a, usv_empty)
 
   # test blockdiagonal
   for i in LinearAlgebra.diagind(blocks(a))
     I = CartesianIndices(blocks(a))[i]
     a[Block(I.I...)] = rand(T, size(blocks(a)[i]))
   end
   usv = svd(a)
-  test_svd(a, usv)
+  @test test_svd(a, usv)
 
   perm = Random.randperm(length(m))
   b = a[Block.(perm), Block.(1:length(n))]
   usv = svd(b)
-  test_svd(b, usv)
+  @test test_svd(b, usv)
 
   # test permuted blockdiagonal with missing row/col
   I_removed = rand(eachblockstoredindex(b))
   c = copy(b)
   delete!(blocks(c).storage, CartesianIndex(Int.(Tuple(I_removed))))
   usv = svd(c)
-  test_svd(c, usv)
+  @test test_svd(c, usv)
 end